Retreat of the Late Pliocene and Lower Pleistocene Crag sea
southwards towards the centre of the syncline. Further downstream the river flowed through the
Vale of St Albans and central Essex into East Anglia, forming a terrace sequence comprising the
Stoke Row, Waterman's Lodge, Westland Green, Satwell, Beaconsfield and Gerrards Cross Members
(Whiteman 1992; Whiteman & Rose 1992; Rose et al. 1999). The correlatives of these in Essex and
Suffolk arc given in Table 1. This route, far to the north of the axis of the London Syncline, might
have been controlled by the Tertiary scarp of the time.
The Kesgrave Formation gravels, as a terrace sequence, have been mapped as far as the Gipping
Valley (Whiteman & Rose 1992; Rose et al. 1999), beyond that Hey (1980) traces them to Norwich
as his 'High Level' Kesgrave Gravels. Hopson & Bridge (1987) disagree with the latter point,
arguing that the Kesgrave Formation is not found in Norfolk and that the Kesgrave Thames did not
cross the alignment of the Bytham River. However, besides Hcy's sites in Norfolk, the Kesgrave
Formation is recognised as a fluvial deposit at Caistor St Edmund, near Norwich (Postma & Hodgson
1988), and is considered to make an input to the shallow marine deposits of the Cromer Stone Bed
at West Runton (Green & McGregor 1990). This is a matter that requires further consideration.
The downstream increase in flint (Table 4) probably indicates that the major source is within the
London Basin, from the Chalk and, probably more importantly, from the Tertiary beds. The quartz/
itc group reduces in number as there is not a significant source within the London Basin to supply or
replenish it. The cherts increase in number, suggesting they are mostly Greensand cherts, coming in
from the Weald. The numbers of acid volcanics is too small to comment on the variation meaningfully
(e.g. 0.8% of a count of 500 would still be only 4 stones), but the very fact that they are there is
important in reconstructing the palaeogeography of the river. Although masked within the generalised
stone counts, in the older members, quartz is more important than quartzite. In later members the
quartz to quartzite ratio moved to unity, indicating a higher input of quartzite, in turn suggesting that
the catchment changed to include the West Midlands more, deriving the quartzite from the
Kidderminster Conglomerate (Fig. 3b).
The 'Sudbury Formation' Thames received tributaries at various points (Fig. 4). South bank tributaries
are indicated by increases in the amount of Lower Greensand chert in the gravels, as well as mapping
of the tributaries in some cases. Those identified are the Lodden-Blackwater, confluent at Reading
(Gibbard 1982, 1985), the Mole-Wey at Ware (Gibbard 1979), Cray-Darent at Ongar (Green et al.
1982) and the Medway in eastern Essex (Bridgland 1988, 1994, 1995). From the evidence of gravel
spreads in eastern Essex, the Medway flowed throughout the period, depositing the Claydons, Daws
Heath and Oakwood Members of the Lower Thames Formation. The Medway members are not
contiguous with the 'Sudbury Formation7 (Fig. 4), hence the correlations are not secure.
The Bytham River may have been confluent with the Thames to the east of Bury St Edmunds as the
gravels arc marked by an increase in quartzite, vein-quartz, Carboniferous chert and Spilsby Sandstone
(Hey 1980; Clarke & Auton 1982; Rose et al. 1999).
The 'Formation' is frequently a gravelly sand rather than a gravel, particularly at its higher levels
where it overlies the Norwich Crag Formation, due to reworking and incorporation of the Crag
sands. The gravels exhibit intra-formational ice-wedge casts, frost cracks and frost-pitted flints,
indicating deposition in periglacial conditions. The braided river sedimentology also supports a
periglacial origin.
There are no interglacial deposits associated with the 'Formation', possibly because the climatic
fluctuations of the late Pliocene/Early Pleistocene were not sufficiently marked to trigger the complex
22
Essex Naturalist (New Series) 18 (2001)